HPC Deployment / Use Cases (EVEREST + DAPHNE: Workshop on Design and Programming High-performance, distributed, reconfigurable and heterogeneous platforms for extreme-scale analytics)

Introduction Challenges HPC Initiatives Deploying DAPHNE CI UGPP MODA Conclusion Reference Aleš Zamuda 7@aleszamuda [DAPHNE] Deployment / Use Cases @ HiPEAC 2023 workshop 1/ 61
High Performance Embedded Architectures and Compilers
(HiPEAC) 2023
EVEREST + DAPHNE: Workshop on
Design and Programming High-performance, distributed, reconﬁgurable
and heterogeneous platforms for extreme-scale analytics
Wednesday, January 18th 2023, 10:00 - 17:30
Argos (Level 1), Pierre Baudis Convention Centre, Toulouse, France
Deployment / Use Cases
Aleš Zamuda <ales.zamuda@um.si>
DAPHNE: Integrated Data Analysis Pipelines for
Large-Scale Data Management, HPC, and Machine Learning
https://daphne-eu.eu/

Introduction & Outline: Aims of this Talk at the
EVEREST + DAPHNE workshop:
1 (5 min) talk a bit about
Use Cases I usually work
• Evolutionary Algorithms
and things like that (ML, AI)
Real examples: science and HPC
2 (5 min) introduce the Vega supercomputer & it’s setup
• because DAPHNE DSL is for High-Performance Computing
3 (3 min) then connect Use Cases to DAPHNE
and explain, how we run DAPHNE in containers
on the EuroHPC infrastructure
• highlighting:
+ the deployment environment
+ core DAPHNE architecture
+ overall overview including applications
4 (2 min) Summary, Q&A

Introduction: Overview (Focus, Use, Scope)
• This contribution focuses on
speeding up of vectorized
benchmarking, which
• includes optimization
algorithms [1].
• These algorithms are used within
Machine Learning (ML)
workloads together with
benchmarking
• in order to compute the
performance of instances of
such algorithms
on a whole benchmark [2].
• Such performance measure
provides a more general
evaluation of an algorithm’s
applicability,
• i.e. an intelligence generality.
• However, the computational
time for whole benchmark
evaluation extends signiﬁcantly
• compared to a single instance
evaluation
• and hence speeding up might
be required
• under time-constrained
conditions, especially when
e.g. used for robotic deep sea
underwater missions [2].
Real examples: science and HPC
[1] Zamuda, A., Lloret, E., Optimizing Data-Driven Models for Summarization as Parallel Tasks. Journal of Computational Science 42, 101101
(2020).
[2] Zamuda, A, Brest, J., Self-adaptive control parameters’ randomization frequency and propagations in differential evolution. Swarm and
Evolutionary Computation 25C, 72-99 (2015).
[3] Zamuda, A., Hernández Sosa, J. D., Success history applied to expert system for underwater glider path planning using differential
evolution. Expert Systems with Applications 119, 155-170 (2019).

Introduction: Vectorized Benchmarking
Opportunities
A closer observation of execution
times for workloads processed in [2] is
provided in Fig. 1, where it is seen that
the execution time (color of the
patches) changes for different
benchmark executions.
Fig. 1: Execution time of full
benchmarks for different instances of
optimization algorithms. Each patch
presents one full benchmark
execution to evaluate an optimization
algorithm.
• Therefore, it is useful to consider speeding up of
benchmarking through vectorization of the tasks that a
benchmark is comprised of.
• These include e.g.,
• parallell data cleaning part of an
individual ML tile [1] or
• synchronization between tasks when
executing parallell geospatial processing
[3].
• To enable the possibilities of data cleaning
(preprocessing) as well as geospatial processing in
parallell, such opportunities ﬁrst need to be found or
designed, if none yet exist for a problem tackled.
• Therefore, this contribution will highlight some
experiences with ﬁnding and designing parallell ML
pipelines for vectorization and observe speedup gained
from that.
• The speeding up focus will be on optimization
algorithms within such ML pipelines, but some
more future work possibilities will also be provided.
[2] Zamuda, A, Brest, J., Self-adaptive control parameters’ randomization frequency and propagations in differential evolution. Swarm and
Evolutionary Computation 25C, 72-99 (2015).
[3] Zamuda, A., Hernández Sosa, J. D., Success history applied to expert system for underwater glider path planning using differential
evolution. Expert Systems with Applications 119, 155-170 (2019).

Challenges
(First part)
Faced 5 types of challenges, leading to the needs to apply HPC
architectures for benchmarking state-of-the-art topics in
1 forest ecosystem modeling, simulation, and
visualization,
2 underwater robotic mission planning,
3 energy production scheduling for hydro-thermal power
plants,
4 understanding evolutionary algorithms, and
5 text summarization.

Challenges 1: Forest Ecosystem Modeling,
Simulation, and Visualization
• HPC need to process spatial data and add procedural
content.
Videos: https://www.youtube.com/watch?list=PL7pmTW8neV7tZf2qx1wV5zbD74sUyHL3B&v=V9YJgYO_sIA

Challenges 2: Underwater Robotic Mission Planning
• Computational Fluid Dynamics (CFD) spatio-temporal model of the
ocean currents for autonomous vehicle navigation path planning.
• Constrained Differential Evolution Optimization
for Underwater Glider Path Planning
in Sub-mesoscale Eddy Sampling.
• Corridor-constrained optimization: eddy
border region sampling — new challenge
for UGPP & DE.
• Feasible path area is constrained —
trajectory in corridor around the border of
an ocean eddy.
The objective of the glider here is to sample the
oceanographic variables more efﬁciently,
while keeping a bounded trajectory.
HPC: develop new methods and evaluate them.
Video: https://www.youtube.com/watch?v=4kCsXAehAmU

Challenges 3: Energy Production Scheduling for
Hydro-thermal Power Plants
A. Glotić, A. Zamuda. Short-term combined economic and emission
hydrothermal optimization by surrogate differential evolution.
Applied Energy, 1 March 2015, vol. 141, pp. 42-56.
DOI: 10.1016/j.apenergy.2014.12.020

Challenges 4: Understanding Evolutionary Algorithms
• Evolutionary algorithms benchmarking to
understand computational intelligence of
these algorithms (→ storage requirement!),
• aim: Machine Learning to design
an optimization algorithm
(learning to learn).
• Example CI Algorithm Mechanism Design:
Control Parameters Self-Adaptation (in DE).
Video: https://www.youtube.com/watch?v=R244LZpZSG0
Application stacks for real code:
inspired by previous computational
optimization competitions in
continuous settings that used
test functions for
optimization application domains:
• single-objective: CEC 2005,
2013, 2014, 2015
• constrained: CEC 2006, CEC 2007, CEC 2010
• multi-modal: CEC 2010, SWEVO 2016
• black-box (target value): BBOB 2009, COCO
2016
• noisy optimization: BBOB 2009
• large-scale: CEC 2008, CEC 2010
• dynamic: CEC 2009, CEC 2014
• real-world: CEC 2011
• computationally expensive: CEC 2013, CEC
2015
• learning-based: CEC 2015
• 100-digit (50% targets): 2019 joined CEC,
SEMCCO, GECCO
• multi-objective: CEC 2002, CEC 2007, CEC
2009, CEC 2014
• bi-objective: CEC 2008
• many objective: CEC 2018
Tuning/ranking/hyperheuristics use. → DEs as usual
winner algorithms.

Challenges 5: Text Summarization
For NLP, part of ”Big Data”.
Terms across sentences are determined
using a semantic analysis using both:
• coreference resolution (using WordNet) and
• a Concept Matrix (from Freeling).
INPUT
CORPUS
NATURAL
LANGUAGE
PROCESSING
ANALYSIS
CONCEPTS
DISTRIBUTION
PER
SENTENCES
MATRIX OF
CONCEPTS
PROCESSING
CORPUS
PREPROCESSING
PHASE
OPTIMIZATION
TASKS
EXECUTION
PHASE
ASSEMBLE
TASK
DESCRIPTION
SUBMIT
TASKS
TO PARALLEL
EXECUTION
OPTIMIZER
+
TASK DATA
ROUGE
EVALUATION
The detailed new method called
CaBiSDETS is developed in the HPC
approach comprising of:
• a version of evolutionary algorithm
(Differential Evolution, DE),
• self-adaptation, binarization,
constraint adjusting, and some more
pre-computation,
• optimizing the inputs to deﬁne the
summarization optimization model.
Aleš Zamuda, Elena Lloret. Optimizing Data-Driven
Models for Summarization as Parallel Tasks. Journal
of Computational Science, 2020, vol. 42, pp. 101101.
DOI: 10.1016/j.jocs.2020.101101

Challenges 6: new DAPHNE Use Cases
Example Use Cases
• DLR Earth Observation
• ESA Sentinel-1/2 datasets  4PB/year
• Training of local climate zone classifiers on So2Sat LCZ42
(15 experts, 400K instances, 10 labels each, ~55GB HDF5)
• ML pipeline: preprocessing,
ResNet-20, climate models
• IFAT Semiconductor Ion Beam Tuning
• KAI Semiconductor Material Degradation
• AVL Vehicle Development Process (ejector geometries, KPIs)
• ML-assisted simulations, data cleaning, augmentation
• Cleaning during exploratory query processing
[So2Sat LC42: https://mediatum.ub.tum.de/1454690]
[Xiao Xiang Zhu et al: So2Sat LCZ42: A
Benchmark Dataset for the Classification of
Global Local Climate Zones. GRSM 8(3) 2020]

EuroHPC Vega

HPC Initiatives
(Second part)
Timeline (as member) of recent impactful HPC initiatives including Slovenia:
• SLING: Slovenian national supercomputing network, 2010-05-03–,
• SIHPC: Slovenian High-Performance Computing Centre, 2016-03-04–
• ImAppNIO: Improving Applicability of Nature-Inspired Optimisation
by Joining Theory and Practice, 2016-03-09–2020-10-31
• cHiPSet: High-Performance Modelling and Simulation for Big Data
Applications, 2015-04-08–2019-04-07,
• HPC RIVR: UPGRADING OF NATIONAL RESEARCH INFRASTRUCTURES,
Investment Program, 2018-03-01–2020-09-15,
• TFoB: IEEE CIS Task Force on Benchmarking, January 2020–,
• EuroCC: National Competence Centres in the framework of EuroHPC,
2020-09-01–(2022-08-31),
• DAPHNE: Integrated Data Analysis Pipelines for Large-Scale Data
Management, HPC, and Machine Learning, 2020-12-01–(2024-11-30).

Initiatives: SLING, SIHPC, HPC RIVR, EuroCC
• There is a federated and orchestrated aim
towards HPC infrastructure in Slovenia,
especially through:
• SLING: Slovenian national supercomputing network
→ has federated the initiative push towards
orchestration of HPC resources across the country.
• SIHPC: Slovenian High-Performance Computing Centre
→ has orchestrated the ﬁrst EU funds application
towards HPC Teaming in the country
(and Participation of Slovenia in PRACE 2).
• HPC RIVR: UPGRADING OF NATIONAL RESEARCH
INFRASTRUCTURES, Investment Program
→ has provided an investment in experimental HPC
infrastructure.
• EuroCC: National Competence Centres in the framework of
EuroHPC
→ has secured a National Competence Centre (EuroHPC).
Vega supercomputer online
Consortium Slovenian High-Performance Computing Centre
Aškerčeva ulica 6
SI-1000 Ljubljana
Slovenia
Ljubljana, 22. 2. 2017
prof. dr. Anwar Osseyran
PRACE Council Chair
Subject: Participation of Slovenia in PRACE 2
Dear professor Osseyran,
In March 2016 a consortium Slovenski superračunalniški center (Slovenian High-Performance
Computing Centre - SIHPC) was established with a founding act (Attachment 1) where article 6
claims that the consortium will join PRACE and that University of Ljubljana, Faculty of
mechanical engineering (ULFME) will represent the consortium in the PRACE. The legal
representative of ULFME in the consortium and therefore, also in the PRACE (i.e., the delegate
in PRACE Council with full authorization) is prof. dr. Jožef Duhovnik, as follows from
appointment of the dean of ULFME (Attachment 2).
The consortium has also agreed that it joins PRACE 2 optional programme as contributing GP.
With best regards,
assist. prof. dr. Aleš Zamuda, vice-president of SIHPC

Initiatives: ImAppNIO, cHiPSet, TFoB, DAPHNE
Aim towards software to run HPC and improve capabilities:
• ImAppNIO: Improving Applicability of Nature-Inspired Optimisation
by Joining Theory and Practice,
→ improve capabilities through benchmarking (to understand (and to
learn to learn)) CI algorithms
• cHiPSet: High-Performance Modelling and Simulation for Big Data
Applications,
→ include HPC in Modelling and Simulation (of the process to be
learned)
• TFoB: IEEE CIS Task Force on Benchmarking,
→ includes CI benchmarking opportunities, where HPC would enable
new capabilities.
• DAPHNE: Integrated Data Analysis Pipelines for Large-Scale Data
Management, HPC, and Machine Learning.
→ to deﬁne and build an open and extensible system infrastructure
for integrated data analysis pipelines, including data management and
processing, high-performance computing (HPC), and machine learning
(ML) training and scoring https://daphne-eu.github.io/

Deploying DAPHNE
(Third part)
MODA (Monitoring and Operational Data Analytics) tools for
• collecting, analyzing, and visualizing
• rich system and application data, and
• my opinion on how one can make sense of the data for
actionable insights.
• Explained through previous examples:
from a HPC User Perspective.

MODA Actionable Insights, Explained From a HPC
User Perspective, Through the Example of
Summarization
Most interesting findings of summarization on HPC example
are
• the fitness of the NLP model keeps increasing with prolonging
the dedicated HPC resources (see below) and that
• the fitness improvement correlates with ROUGE evaluation in
the benchmark, i.e. better summaries.
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
1 10 100 1000 10000
ROUGE-1R
ROUGE-2R
ROUGE-LR
ROUGE-SU4R
Fitness (scaled)
Hence, the use of HPC
significantly
contributes to
capability of this NLP
challenge.
However, the MODA insight also provided the
useful task running times and resource
usage.

Running the Tasks on HPC: ARC Job Preparation
Parallel summarization tasks on grid through ARC.

Running the Tasks on HPC: ARC Job Submission,
Results Retrieval & Merging [JoCS2020]
Through an HPC
approach and by
parallelization of tasks,
a data-driven
summarization model
optimization yields
improved benchmark
metric results (drawn
using gnuplot merge).
MODA is needed
to run again and
improve upon, to
forecast how to
set required task
running time and
resources
(predicting system
response).

Monitoring and Operational Data Analytics
• Monitor used (jobs, CPU/wall time, etc.):
Smirnova, Oxana. The Grid Monitor. Usage manual,
Tech. Rep. NORDUGRID-MANUAL-5, The NorduGrid
Collaboration, 2003.
http://www.nordugrid.org/documents/
http://www.nordugrid.org/manuals.html
http://www.nordugrid.org/documents/monitor.pdf
• Deployed at:
www.nordugrid.org/monitor/
• NorduGrid Grid Monitor
Sampled: 2021-06-28 at 17-57-08
• Nation-wide in Slovenia:
https://www.sling.si/gridmonitor/loadmon.php
http://www.nordugrid.org/monitor/index.php?
display=vo=Slovenia

MODA Example From: ARC at Jost
Example experiments from DOI: 10.1016/j.swevo.2015.10.007 (SPSRDEMMS)
– job YYULDmGOXpmnmmR0Xox1SiGmABFKDmABFKDmrtMKDmABFKDm66faPo.
Sample ARC ﬁle gridlog/diag (2–3 day Wall Times).
runtimeenvironments=APPS/ARNES/MPI−1.6−R ;
CPUUsage=99%
MaxResidentMemory=5824kB
AverageUnsharedMemory=0kB
AverageUnsharedStack=0kB
AverageSharedMemory=0kB
PageSize=4096B
MajorPageFaults=4
MinorPageFaults=1213758
Swaps=0
ForcedSwitches=36371494
WaitSwitches=170435
Inputs=45608
Outputs=477168
SocketReceived=0
SocketSent=0
Signals =0
nodename=wn003 . arnes . s i
WallTime=148332s
Processors=16
UserTime=147921.14s
KernelTime =2.54 s
AverageTotalMemory=0kB
AverageResidentMemory=0kB
LRMSStartTime=20150906104626Z
LRMSEndTime=20150908035838Z
exitcode=0

EuroCC HPC: Vega (TOP500 #106, HPCG #56 — June 2021)
• Researchers apply to EuroHPC JU calls for access.
• Regular calls opened in 2021 fall (Benchmark & Development).
• https://prace-ri.eu/benchmark-and-development-access-information-for-applicants/
• 60% capacities for national share (70% OA, 20% commercial, 10% host (community, urgent
priority of national importance, maintenance)) + 35% EuroHPC JU share (approved applications)
• Has a SLURM dev partition for SSH login (SLURM partitions w/ CPUs: login[0001-0004]=128;
login[0005-0008]=64; cn[0001-384,0577-0960]=256; cn[0385-0576]=256; gn[01-60]=256).
Listing 1: Setting up at Vega — slurm dev partition access (login).
1 [ ales . zamuda@vglogin0007 ˜]$ s i n g u l a r i t y pull qmake . s i f docker : / / ak352 /qmake−opencv
2 [ ales . zamuda@vglogin0007 ˜]$ s i n g u l a r i t y run qmake . s i f bash
3 cd sum; qmake ; make clean ; make
4
5 [ ales . zamuda@vglogin0007 ˜]$ cat runme . sh
6 # ! / bin / bash
7 cd sum && time mpirun
8 −
−mca btl openib warn no device params found 0
9 . / summarizer
10 −
−useBinaryDEMPI −
−i n p u t f i l e mRNA−1273−t x t
11 −
−withoutStatementMarkersInput
12 −
−printPreprocessProgress calcInverseTermFrequencyndTermWeights
13 −
−printOptimizationBestInGeneration
14 −
−summarylength 600 −
−NP 200
15 −
−GMAX 400
16 > summarizer . out . $SLURM PROCID
17 2> summarizer . err . $SLURM PROCID
-0.65
-0.6
-0.55
-0.5
-0.45
-0.4
-0.35
-0.3
-0.25
-0.2
-0.15
-0.1
1 10 100
Evaluation

MODA at First EuroCC HPC Vega Supercomputer
Listing 2: Runnig at Vega & MODA.
1 ===================================================================== GMAX=200 =====
2 [ ales . zamuda@vglogin0002 ˜]$ srun −
−cpu−bind=cores −
−nodes=1 −
−ntasks−per−node=101
3 −
−cpus−per−task=2 −
−
mem=180G s i n g u l a r i t y run qmake . s i f bash
4 srun : job 4531374 queued and waiting for resources
5 srun : job 4531374 has been allocated resources
6 [ ”$SLURM PROCID” = 0 ] && . / runme . sh
7 real 5m22.475 s
8 user 484m42.262 s
9 sys 1m38.304 s
10 ===================================================================== NODES=51 =====
−nodes=1 −
12 −
−
16 real 13m57.851 s
17 user 431m25.833 s
18 sys 0m29.272 s
19 ===================================================================== GMAX=400 =====
−nodes=1 −
21 −
−
25 real 6m14.687 s
26 user 590m45.641 s
27 sys 1m40.930 s

More Output: SLURM accounting
Listing 3: Example accounting tool at Vega: sacct.
[ ales . zamuda@vglogin0002 ˜]$ sacct
4531374. ext+ extern vega−users 202 COMPLETED 0:0
[ ales . zamuda@vglogin0002 ˜]$ sacct −j 4531374 −j 4531746 −j 4532697
−o MaxRSS , MaxVMSize , AvePages
MaxRSS MaxVMSize AvePages
−
−
−
−
−
−
−
−
−
− −
−
−
−
−
−
−
−
−
− −
−
−
−
−
−
−
−
−
−
0 217052K 0
26403828K 1264384K 22
0 217052K 0
13325268K 1264380K 0
0 217052K 0
26404356K 1264384K 30
Future MODA testings:
• testing the web interface for job analysis (as available from HPC RIVR);
• proﬁling MPI inter-node communication;
• use proﬁlers and monitoring tools available
— in the context of heterogeneous setups, like e.g.
• TAU Performance System — http://www.cs.uoregon.edu/research/tau/home.php,
• LIKWID Performance Tools — https://hpc.fau.de/research/tools/likwid/.

Deploying DAPHNE on Vega
Main documentation ﬁle:
Deploy.md

SLURM

More HPC User Perspective Nation-wide in Slovenia
More: at University of Maribor, Bologna study courses for
teaching (training) of Computer Science at cycles — click URL:
• level 1 (BSc)
• year 1: Programming I – e.g. C++ syntax
• year 2: Computer Architectures – e.g. assembly, microcode, ILP
• year 3: Parallell and Distributed Computing – e.g. OpenMP, MPI, CUDA
• level 2 (MSc)
• year 1: Cloud Computing Deployment and Management – e.g. arc, slurm,
Hadoop, containers (docker, singularity) through
virtualization
• level 3 (PhD)
• EU and other national projects research:
HPC RIVR, EuroCC, DAPHNE, ... – e.g. scaling new systems
of CI & Operational Research of ... over HPC
• IEEE Computational Intelligence Task Force on Benchmarking
• Scientiﬁc Journals (e.g. SWEVO, TEVC, JoCS, ASOC, INS)
These contribute towards Sustainable Development of HPC.

Background (more):
ML workloads

Background on HPC ML Workloads:
Multi-application domain generalized Computational
Intelligence (CI) through HPC
• With the ubiquity of HPC and Cloud Computing,
• connectible using versatile interfaces for their utilisation,
• example successes of these systems is computational
intelligence.
• One of designs for these services is:
• using evolutionary optimization approaches,
• needing much efﬁciently parallelizable data processing
power.
• There are several application domains of this approach:
• generalized numerical functions problems and
• other parallel real world problems, such as
• text processing,
• molecular modelling,
• evolutionary computer vision, and
• robotics.

CI Algorithm Design: Control Parameters
Self-Adaptation
• Through more suitable values of control parameters the
search process exhibits a better convergence (CI —
Computational Intelligence),
• therefore the search converges faster to better solutions,
which survive with greater probability and they create
more offspring and propagate their control parameters
• Recent study with cca. 10 million runs of SPSRDEMMS:
A. Zamuda, J. Brest. Self-adaptive control parameters’
randomization frequency and propagations in differential
evolution. Swarm and Evolutionary Computation, 2015,
vol. 25C, pp. 72-99.
DOI 10.1016/j.swevo.2015.10.007.
– SWEVO 2015 RAMONA / SNIP 5.220
• Deployed using arcsub.
• Update: DISH algorithm —
https://doi.org/10.1016/j.swevo.2018.10.013
(Outperforms on CEC 2017 functions.)

Existing Challenges: Benchmarks for Known Environments
Inspired by previous computational optimization competitions in continuous
settings that used test functions for optimization application domains:
• single-objective: CEC 2005, 2013, 2014, 2015
• constrained: CEC 2006, CEC 2007, CEC 2010
• multi-modal: CEC 2010, SWEVO 2016
• black-box (target value): BBOB 2009, COCO 2016
• noisy optimization: BBOB 2009
• large-scale: CEC 2008, CEC 2010
• dynamic: CEC 2009, CEC 2014
• real-world: CEC 2011
• computationally expensive: CEC 2013, CEC 2015
• learning-based: CEC 2015
• 100-digit (50% targets): 2019 joined CEC, SEMCCO, GECCO
• multi-objective: CEC 2002, CEC 2007, CEC 2009, CEC 2014
• bi-objective: CEC 2008
• many objective: CEC 2018
Tuning/ranking/hyperheuristics use. → DEs as usual winner
algorithms.

IEEE Congress on Evolutionary Computation (CEC)
Competitions (1/4)
• Storn, Rainer, and Kenneth V. Price. ”Minimizing the Real Functions of the ICEC’96 Contest by
Differential Evolution.” International Conference on Evolutionary Computation. 1996.
• ...
• CEC 2005 Special Session / Competition on
Evolutionary Real Parameter single objective optimization
Evolutionary Constrained Real Parameter single objective optimization
Performance Assessment of real-parameter MOEAs
large scale single objective global optimization with bound constraints
• CEC 2008
Scale-Invariant Optimisation Competition ”Mountains or Molehills”
Dynamic Optimization (Primarily composition functions were used)
Performance Assessment of real-parameter MOEAs

Competitions (2/4)
large-scale single objective global optimization with bound constraints
Evolutionary Constrained Real Parameter single objective optimization
• CEC 2010 Special Session on
Niching Introduces novel scalable test problems
• CEC 2011 Competition on Testing Evolutionary Algorithms on Real-world
Numerical Optimization Problems
Real Parameter Single Objective Optimization
Real Parameter Single Objective Optimization
(incorporates expensive functions)
• CEC 2014: Dynamic MOEA Benchmark Problems

Competitions (3/4)
Real Parameter Single Objective Optimization (incorporates 3
scenarios)
• CEC 2015 Black Box Optimization Competition
• CEC 2015 Dynamic Multi-Objective Optimization
• CEC 2015 Optimization of Big Data
• CEC 2015 Large Scale Global Optimization
• CEC 2015
Bound Constrained Single-Objective Numerical Optimization
• CEC 2015
Optimisation of Problems with Multiple Interdependent Components
• CEC 2015 Niching Methods for Multimodal Optimization

Competitions (4/4)
Real Parameter Single Objective Optimization (incorporates 4
scenarios)
• CEC 2016 Big Optimization (BigOpt2016)
• CEC 2016 Niching Methods for Multimodal Optimization
• CEC 2016 Special Session Associated with Competition on
Bound Constrained Single Objective Numerical Optimization
• CEC2017 Special Session / Competition on Real Parameter Single
Objective Optimization (also constrained)
• CEC2018 Special Session / Competition on Real Parameter Single
Objective Optimization
• CEC2019 Special Session / Competition on 100-Digit Challenge on
Single Objective Numerical Optimization
• CEC 2020, CEC 2021, 2022, ...: see IEEE Task Force on Benchmarking
(https://cmte.ieee.org/cis-benchmarking/) listings (incl. IEEE
TEVC, MIT EC, MDPI SIs, GitHub, etc.)

Some More Benchmark Function Sets
The Genetic and Evolutionary Computation Conference (GECCO):
• GECCO 2014 Windfarm Layout Optimization Competition
• GECCO 2014
Permutation-based Combinatorial Optimization Problems
• GECCO 2015 Black Box Optimization (BBComp)
• GECCO 2015 Combinatorial Black-Box Optimization (CBBOC)
• GECCO 2019 Workshop & Competition on Numerical Optimization
• GECCO 2022 (Boston) Open Optimization Competition 2022: Better
Benchmarking of Sampling-Based Optimization Algorithms
• GECCO 2022 (Boston) SpOC: Space Optimisation Competition
• Swarm and Evolutionary Computation: Novel benchmark functions
for continuous multimodal optimization with comparative results
(2015)
• Benchmarks for natural architecture design:
Spatial Tree Morphology Reconstruction
(seeded/pre-processed/vectorized/multi-objective)
• Benchmarks for ocean exploration with underwater robots:
Underwater Glider Path Planning
(unconstraint/constraint/variable/multi-objective)

Real World Industry Challenges: Motivation
• Optimization of Real World Industry Challenges (RWIC),
• selected as CEC 2011 Real World Optimization Problems,
• a benchmark set contains functions modelling the
problems,
• assessment on all 22 functions of CEC 2011 set,
• functions with constraints are handled with additional
care.

Selected Challenges in the CEC 2011 Benchmark (1/2)
• Decomposition of radio FM signals,
• determination of ternary protein structure
• Lennard-Jones inter-atom energy potential,
• parameterization of a chemical process,
• methylcyclopentane → benzene,
• control parameterization for chemical reaction in a
continuous stirred tank reactor,
• inter-atom potential in covalent Silicon systems
• Tersoff energy potential,
• radar spectrum signal broadcast parameterization,
• spread spectrum radar polly phase code design,
• electrical transmission network expansion planning,
• new lines for transmission selection.

Selected Challenges in the CEC 2011 Benchmark (2/2)
• Large scale transmission pricing,
• circular antenna array design,
• dynamic economic dispatch (with power generator
control),
• static economic load dispatch (of power generated),
• hydrothermal scheduling (among hydro/thermal units),
• spacecraft trajectory optimization,
• Mercury (Messenger),
• Saturn (Cassini).
• The collection includes 22 functions, with constraints as:
1 non-feasible evaluation is NaN when constraints are not
met, or
2 the constraints are included in the function evaluation
value.

HPC Application 2:
Underwater Glider: Autonomous, Unmanned, Robotic
• underwater glider – navigating sea oceans,
• Autonomous Underwater Vehicle (AUV)
̸=
Unmanned Aerial Vehicle (UAV)
• AUV Slocum model (expertise in domain of ULPGC, work
with J. D. Hernández Sosa)
Images:
”Photo: Richard Watt/MOD” (License: OGL v1.0)
Slocum-Glider-Auvpicture 5.jpg (License: Public Domain)
MiniU.jpg (License: CC-BY-SA 3.0)

The Buoyancy Drive and Submarine Probes
Usefulness
• Driving ”yoyo” uses little energy, most only on descent
and rise (pump); also for maintaining direction little
power is consumed.
+ Use: improving ocean models with real data,
+ the real data at the point of capture,
+ sampling ﬂow of oil discharges,
+ monitoring cable lines, and
+ real-time monitoring of different
sensor data.
1
http://spectrum.ieee.org/image/1523708

Preparations – Simulation Scenarios
https://www.google.si/maps/@28.059806,-15.998355,650054m/data=!3m1!1e3

Trajectory Optimization: P201,ESTOC2013 3
+ BigData, MyOcean IBI,
satelite link, GPS location
The real trajectory and collected data is available in a Google Earth KML ﬁle at the EGO network:
http://www.ego-network.org/dokuwiki/doku.php?glider=P201,ESTOC2013_3

Constrained Differential Evolution Optimization
for Underwater Glider Path Planning
in Sub-mesoscale Eddy Sampling
• Corridor-constrained optimization:
eddy border region sampling
• new challenge for UGPP & DE
• Feasible path area is constrained
• trajectory in corridor around
the border of an ocean eddy
The objective of the glider here is to
sample the oceanographic variables
more efﬁciently,
while keeping a bounded trajectory

HoP — New Trajectories:
Success history applied to expert system for
underwater glider path planning using differential
evolution
• Improved underwater glider path
planning mission scenarios:
optimization with L-SHADE.
• Several conﬁgured algorithms
are also compared to,
analysed, and further
improved.
• Outranked all other previous
results from literature and
ranked ﬁrst in comparison.
• New algorithm yielded
practically stable and
competitive output
trajectories.
•

Ranking UGPP —
Benchmarking
Aggregation
• Statistically,
all results
from previous paper
were outperformed.
• Main reasons:
tuning (NP),
parameter control
(L-SHADE).

Data Analytics of the Operational Data Monitoring
(MODA)
• MODA: data obtained from individual’s task gridlog/diag
WallTime.
for i in *; do [ −f ” $i / gridlog / diag ” ] || continue ;
echo −n $ ( cat $i /CONFIG . sh | xargs ) ” ”
grep WallTime $i / gridlog / diag
done 2>/dev / null |
sed −e ’ s / export / / g ’ −e ’ s /=/ /g ’ |
grep WallTime | grep tauF |
sed −e ’ s / tauF 0 . . . tauCR 0 . . . t / t /g ’ |
sed −e ’ s / s$ / / g ’ | sort −g
• MODA: data obtained from individual’s task gridlog/diag
WallTime.
gnuplot <EOS
set t i t l e ”MODA( Wall Time [ in seconds ] ) of SPSRDEMMS for different F , CR
( both between 0.0 and 1 . 0 ) ”
set palette defined (0 0 0 0.5 , 1 0 0 1 , 2 0 0.5 1 , 3 0 1 1 , 4 0.5 1 0.5 ,
5 1 1 0 , 6 1 0.5 0 , 7 1 0 0 , 8 0.5 0 0)
unset z t i c s
set y t i c s
set x t i c s
splot ’moda’ u : 2 : 4 : 6 with points pt 5 ps 9 palette t ” ”
EOS

Data Analytics of the Operational Data Monitoring
(MODA): Example Results
τF τCR Time [s]
0.70 0.12 6167
0.36 0.18 13904
0.10 0.25 15426
0.25 0.90 15643
0.20 0.05 16039
0.05 0.85 16128
0.05 0.05 16586
0.05 1.00 16902
0.15 0.65 17207
0.05 0.60 17269
... ... ...
0.2 0.6 183938
0.95 0.60 189802
0.72 0.18 259200
0.77 0.18 259203
0.77 0.12 259211
0.73 0.14 259214
0.79 0.16 259217
0.78 0.15 259222
0.78 0.19 259226
0.73 0.12 259229
Total combined WallTime time: 1.03658e+08
seconds (≈ 3.2 years).
In practice, when Vectorized: ≈ 1 month on
HPC (jobs 21 Aug 2015 – 24 Sep 2015 ≈ 35
days).
• MODA: data obtained from
individual’s task gridlog/diag
WallTime.
• ML task MODA plotted.
• Speedup: 35 days vs. 3.2 years.
• Capability gained: paper
revision.
• Gain impact: reference.

Conclusion

Conclusion
Summary: vectorized ML benchmarking, Vega, deployment.
Thanks!
Acknowledgement: this work is supported by DAPHNE, funded by
the European Union’s Horizon 2020 research and innovation
programme under grant agreement No 957407; and ARRS
programme P2-0041.
Questions?

Biography and References: Organizations
• Associate Professor at University of Maribor, Slovenia
• Continuous research programme funded by Slovenian Research Agency,
P2-0041: Computer Systems, Methodologies, and Intelligent Services
• Associate Editor: Swarm and Evolutionary Computation
• IEEE (Institute of Electrical and Electronics Engineers)
senior
• IEEE Computational Intelligence Society (CIS), senior member
• IEEE CIS Task Force on Benchmarking, chair Website link
• IEEE CIS, Slovenia Section Chapter (CH08873), chair
• IEEE Slovenia Section, 2018–2021 vice chair
• IEEE Young Professionals Slovenia, past chair
• ACM SIGEVO (Special Interest Group on Genetic and Evolutionary Computation); EurAI; SLAIS
• Co-operation in Science and Techology (COST) Association Management
Committee, member:
• CA COST Action CA15140: Improving Applicability of Nature-Inspired
Optimisation by Joining Theory and Practice (ImAppNIO), WG3 VC
• ICT COST Action IC1406 High-Performance Modelling and Simulation for
Big Data Applications (cHiPSet); SI-HPC; HPC-RIVR user
• EU H2020 Research and Innovation project, holder for UM part: Integrated
Data Analysis Pipelines for Large-Scale Data Management, HPC, and
Machine Learning (DAPHNE),
https://cordis.europa.eu/project/id/957407

Biography and References: Top Publications
• Aleš Zamuda, Elena Lloret. Optimizing Data-Driven Models for Summarization as Parallel Tasks.
Journal of Computational Science, 2020, vol. 42, pp. 101101. DOI 10.1016/j.jocs.2020.101101.
• A. Zamuda, J. D. Hernández Sosa. Success history applied to expert system for underwater glider path
planning using differential evolution. Expert Systems with Applications, 2019, vol. 119, pp. 155-170. DOI
10.1016/j.eswa.2018.10.048
• A. Viktorin, R. Senkerik, M. Pluhacek, T. Kadavy, A. Zamuda. Distance Based Parameter Adaptation for
Success-History based Differential Evolution. Swarm and Evolutionary Computation, 2019, vol. 50, pp.
100462. DOI 10.1016/j.swevo.2018.10.013.
• A. Zamuda, J. Brest. Self-adaptive control parameters’ randomization frequency and propagations
in differential evolution. Swarm and Evolutionary Computation, 2015, vol. 25C, pp. 72-99.
DOI 10.1016/j.swevo.2015.10.007.
• A. Zamuda, J. D. Hernández Sosa, L. Adler. Constrained Differential Evolution Optimization for
Underwater Glider Path Planning in Sub-mesoscale Eddy Sampling. Applied Soft Computing, 2016,
vol. 42, pp. 93-118. DOI 10.1016/j.asoc.2016.01.038.
• A. Zamuda, J. D. Hernández Sosa. Differential Evolution and Underwater Glider Path Planning
Applied to the Short-Term Opportunistic Sampling of Dynamic Mesoscale Ocean Structures.
Applied Soft Computing, vol. 24, November 2014, pp. 95-108. DOI 10.1016/j.asoc.2014.06.048.
• A. Zamuda, J. Brest. Vectorized Procedural Models for Animated Trees Reconstruction using
Differential Evolution. Information Sciences, vol. 278, pp. 1-21, 2014. DOI 10.1016/j.ins.2014.04.037.
• A. Zamuda, J. Brest. Environmental Framework to Visualize Emergent Artiﬁcial Forest Ecosystems.
Information Sciences, vol. 220, pp. 522-540, 2013. DOI 10.1016/j.ins.2012.07.031.
• A. Glotić, A. Zamuda. Short-term combined economic and emission hydrothermal optimization by
surrogate differential evolution. Applied Energy, 1 March 2015, vol. 141, pp. 42-56.
DOI 10.1016/j.apenergy.2014.12.020.
• H. Hamann, Y. Khaluf, J. Botev, M. Divband Soorati, E. Ferrante, O. Kosak, J.-M. Montanier, S. Mostaghim,
R. Redpath, J. Timmis, F. Veenstra, M. Wahby and A. Zamuda. Hybrid Societies: Challenges and
Perspectives in the Design of Collective Behavior in Self-organizing Systems. Frontiers in Robotics
and AI, 2016, vol. 3, no. 14. DOI 10.3389/frobt.2016.00014.
• J. Šilc, A. Zamuda. Special Issue on ”Bioinspired Optimization” (guest editors). Informatica - An
International Journal of Computing and Informatics, 2015, vol. 39, no. 2, pp. 1-122.

Biography and References: Bound Speciﬁc to HPC
PROJECTS:
• DAPHNE: Integrated Data Analysis Pipelines for Large-Scale Data Management,
HPC, and Machine Learning
• ICT COST Action IC1406 High-Performance Modelling and Simulation for Big
Data Applications
• SLING: Slovenian national supercomputing network
• SI-HPC: Slovenian corsortium for High-Performance Computing
• UM HPC-RIVR: Supercomputer at UM, https://www.hpc-rivr.si/
• SmartVillages: Smart digital transformation of villages in the Alpine Space
• Interreg Alpine Space,
https://www.alpine-space.eu/projects/smartvillages/en/home
• Interactive multimedia digital signage (PKP, Adin DS)
EDITOR:
• SWEVO (Top Journal), Associate Editor
• Mathematics-MDPI, Special Issue: Evolutionary Algorithms in Engineering Design Optimization
• Journal of advanced engineering and computation (member of editorial board since 2019). Viet Nam: Ton
Duc Thang University, 2017-. ISSN 2588-123X.
• Cloud Computing and Data Science (Associate Editor, since 2019). Universal Wiser Publisher Pte.Ltd.
• D. Gleich, P. Planinšič, A. Zamuda. 2018 25th International Conference on Systems, Signals and Image
Processing (IWSSIP). IEEE Xplore, Maribor, 20-22 June 2018.
• General Chair: 7-th Joint International Conferences on Swarm, Evolutionary and Memetic Computing
Conference (SEMCCO 2019) & Fuzzy And Neural Computing Conference (FANCCO 2019), Maribor, Slovenia,
EU, 10-12 July 2019; co-editors: Aleš Zamuda, Swagatam Das, Ponnuthurai Nagaratnam Suganthan, Bijaya
Ketan Panigrahi.

Biography and References: More on HPC
RESEARCH PUBLICATIONS:
• Aleš Zamuda, Elena Lloret. Optimizing Data-Driven Models for Summarization as Parallel Tasks.
Journal of Computational Science, 2020, vol. 42, pp. 101101. DOI 10.1016/j.jocs.2020.101101.
• Aleš Zamuda, Vincenzo Crescimanna, Juan C. Burguillo, Joana Matos Dias, Katarzyna Wegrzyn-Wolska,
Imen Rached, Horacio González-Vélez, Roman Senkerik, Claudia Pop, Tudor Cioara, Ioan Salomie, Andrea
Bracciali. Forecasting Cryptocurrency Value by Sentiment Analysis: An HPC-Oriented Survey of the
State-of-the-Art in the Cloud Era. Kołodziej J., González-Vélez H. (eds) High-Performance Modelling and
Simulation for Big Data Applications. Lecture Notes in Computer Science, vol 11400, 2019, pp. 325-349.
DOI 10.1007/978-3-030-16272-6 12.
• Nenad Karolija, Aleš Zamuda. On cloud-supported web-based integrated development environment
for programming dataﬂow architectures. MILUTINOVIĆ, Veljko (ur.), KOTLAR, Milos (ur.). Exploring the
DataFlow supercomputing paradigm: example algorithms for selected applications, (Computer
communications and networks (Internet), ISSN 2197-8433), 2019, pp. 41-51. DOI
10.1007/978-3-030-13803-5 2.
• Simone Spolaor, Marco Gribaudo, Mauro Iacono, Tomas Kadavy, Zuzana Komı́nková Oplatková, Giancarlo
Mauri, Sabri Pllana, Roman Senkerik, Natalija Stojanovic, Esko Turunen, Adam Viktorin, Salvatore Vitabile,
Aleš Zamuda, Marco S. Nobile. Towards Human Cell Simulation. Kołodziej J., González-Vélez H. (eds)
High-Performance Modelling and Simulation for Big Data Applications. Lecture Notes in Computer
Science, vol 11400, 2019, pp. 221-249. DOI 10.1007/978-3-030-16272-6 8.
• A. Zamuda, J. D. Hernandez Sosa, L. Adler. Improving Constrained Glider Trajectories for Ocean Eddy
Border Sampling within Extended Mission Planning Time. IEEE Congress on Evolutionary Computation
(CEC) 2016, 2016, pp. 1727-1734.
• A. Zamuda. Function evaluations upto 1e+12 and large population sizes assessed in distance-based
success history differential evolution for 100-digit challenge and numerical optimization scenarios
(DISHchain 1e+12): a competition entry for ”100-digit challenge, and four other numerical optimization
competitions” at the genetic and evolutionary computation conference (GECCO) 2019. Proceedings of the
Genetic and Evolutionary Computation Conference Companion (GECCO 2019), 2019, pp. 11-12.
• ... several more experiments for papers run using HPCs.
• ... also, pedagogic materials in Slovenian and English — see Conclusion .

Promo materials: Calls for Papers, Informational
Websites
CS FERI WWW
CIS TFoB
CFPs WWW
LinkedIn
Twitter

HPC Deployment / Use Cases (EVEREST + DAPHNE: Workshop on Design and Programming High-performance, distributed, reconfigurable and heterogeneous platforms for extreme-scale analytics)

Recommended

Recommended

More Related Content

Similar to HPC Deployment / Use Cases (EVEREST + DAPHNE: Workshop on Design and Programming High-performance, distributed, reconfigurable and heterogeneous platforms for extreme-scale analytics)

Similar to HPC Deployment / Use Cases (EVEREST + DAPHNE: Workshop on Design and Programming High-performance, distributed, reconfigurable and heterogeneous platforms for extreme-scale analytics) (20)

More from University of Maribor

More from University of Maribor (17)

Recently uploaded

Recently uploaded (20)

HPC Deployment / Use Cases (EVEREST + DAPHNE: Workshop on Design and Programming High-performance, distributed, reconfigurable and heterogeneous platforms for extreme-scale analytics)